- Ben Merva
What to ask before choosing a partner and an RFP template for scoring vendors.
Table of Contents
The Category
What is a commerce intelligence system?
A commerce intelligence system is the operational source of truth for how your brand sells through every retailer. It runs on retail-direct data, the feeds that come straight from each retailer and reconcile with what you ship and get paid for. It resolves those fragmented feeds into one clean view at the SKU-and-location level, then builds the metrics your teams run on (POS, inventory, weeks of supply, forecast accuracy), and helps you act on them to recover sales, protect margin, and free up working capital.
Start Here
Most guides list features. This one gives you the questions.
Each question below comes with why it matters and what a weak answer sounds like, so you can tell a real capability from a good demo. Use it with any vendor you’re evaluating, us included. The RFP template at the end turns it into a scorecard.
THE SEVEN QUESTIONS THAT SEPARATE PLATFORMS
What to ask, why it matters, what a weak answer sounds like
Download the PDF version of this guide to view an example and complete your worksheet.
1 · Coverage
How much do you actually connect, and who maintains it?
Ask: How many retailers, distributors, e-commerce, social commerce, online marketplaces, and 3PL sources do you connect, and who builds and maintains each connection?
Why it matters: Your distribution keeps growing. If coverage stops at your top accounts, the long tail becomes a blind spot, and maintaining the connections becomes your team’s second job.
Weak answer: “We connect the major retailers.” A number they won’t put in writing.
2 · Normalization
Can you connect insights for the same product across different data sources?
Ask: Do you resolve the same product to one master item across every retailer automatically, and deliver it at the SKU and store level, not just a national roll-up?
Why it matters: Lining up columns isn’t lining up products, and a national number hides the store where you’re out of stock. You act at the SKU-location level, so your data has to live there.
Weak answer: A mapping file you maintain, or data only at the retailer or region level.
3 · Depth
Does it stop at the shelf?
Ask: Do you connect past POS data to DC, ERP, and 3PL data, or only retailer sell-through?
Why it matters: The shelf is half the picture. Availability, replenishment, and margin problems live between the store and your own systems.
Weak answer: POS and e-commerce only, with no path to your internal data.
4 · AI & AI Agents
What's usable today, and who's in control?
Ask: What agentic AI is live today (not on the roadmap)? Do agents monitor your full network or a subset of major retailers? What is the review, edit, and approval step before anything is sent?
Why it matters: Everyone says “AI” now. What matters is whether it acts across your whole business and whether a human stays in control.
Weak answer: “Coming soon,” major-retailers-only, or agents that act without an approval step.
5 · Data Integrity
Is the data fresh, complete, and trustworthy?
Ask: How fresh is the data, is it retailer-direct, and how do you handle failed loads, missing files, and restatements?
Why it matters: A number nobody trusts gets re-checked by hand. Silent gaps cost you the decision.
Weak answer: Weekly or estimated data, and no monitoring for missing files.
6 · Time To Value
How fast are you actually live?
Ask: How fast is our priority feed live? How long for a full rollout? Who does the build, you or us?
Why it matters: A platform that takes two quarters to show value is a data project that can stall before it ships.
Weak answer: A months-long timeline, or a build that lands on your team.
7 · Ownership, Security, & Data Use
Is your data yours, and is it governed?
Ask: Do we own our data, and can we export it at any time to our data warehouse, BI and analytics tools, ERP and planning systems, and Excel or Slides? Are you SOC 2 Type II with encryption and an isolated instance, and is our data ever used to train models?
Why it matters: A platform you can’t easily get your data out of, or that trains its models on it, is a lock-in and a governance risk.
Weak answer: Limited export, data you can’t fully take with you, or data that trains shared models.
Since You'll Ask
How Alloy.ai answers these
- Coverage. 500+ retailer, distributor, e-commerce, social commerce, online marketplace, and 3PL connections, built and maintained by Alloy, self-healing through portal changes. We can also build custom integrations for any source type listed if you don’t find what you need in our pre-built integrations.
- Normalization. The same item resolved to one master across every commerce partner, at the SKU and location level, automatically where possible.
- Depth. Beyond POS and into DC, ERP, and 3PL data, so the shelf connects to your operation.
- AI & Agents. Replenishment and Performance Reporting agents live today (with more on the roadmap), plus Lens, an AI analyst, for asking questions in plain language, no SQL. Agents draft the work; a human approves, edits, or rejects before anything is sent. Replenishment drafts orders for every partner where the data supports it (store-level sales, DC inventory, store-to-DC mapping), expanding as that data becomes available. Lens can answer complex analytics questions, build reports, act as a strategic thought partner, and surface relevant insights from your data.
- Time to value. Live data in 3 business days; a full 10+ retailer rollout in about 8 weeks, and Alloy does the build.
- Security. SOC 2, encryption at rest and in transit, isolated per-customer instance. Your data never trains our models or anyone else’s.
- Ownership. Your data stays yours. Automate data sharing or export it at any time to your data warehouse, BI and analytics tools, ERP and planning systems, or Microsoft Office / Google Suite. No lock-in.
TRUSTED BY THE WORLD’S TOP BRANDS
Book a Demo
See Alloy.ai
Synchronize execution to eliminate waste, mitigate risk, and capture every revenue opportunity.
Once you submit a demo request form:
- A sales rep will email you within 24 hours to schedule a brief intro call.
- Our team will provide an overview of Alloy.ai and learn more about your data goals and objectives.
- We’ll create and walk you through a live demo, customized to your specific data needs.
Schedule a Demo
The RFP Template
Score every vendor on the same bar
Score each requirement 0, 3, or 5. Average the scores in a category, apply the weight, and sum them to a total out of 100. The Alloy column is pre-filled; the three blank columns are for the other vendors you evaluate.
Scoring Scale
- 5 = meets fully, out of the box
- 3 = partial, or needs configuration
- 0 = no, or roadmap
Category Weights
| Category | Weight |
|---|---|
| Coverage & integration | 20% |
| AI & agents | 20% |
| Normalization | 15% |
| Depth: supply chain | 15% |
| Data integrity | 10% |
| Forecasting & analytics | 8% |
| Ownership, security & support | 12% |
| Total | 100% |
Requirements: Coverage and Integration
| Requirement | Alloy.ai | Vendor 2 | Vendor 3 | Vendor 4 |
|---|---|---|---|---|
| How many retailer, distributor, e-comm, social commerce, online marketplace, and 3PL connections, and who maintains them? | 500+ integrations, built and maintained by Alloy (self-healing) plus custom integrations can be built as needed | |||
| Connects e-commerce, online marketplaces, and social commerce, not just brick-and-mortar POS? | Yes | |||
| Unifies POS, e-comm, PO, shipment, receipt, and owned-inventory into one model, at SKU and location level? | Yes, daily | |||
| Can write data into our warehouse (Snowflake, Redshift, GCP)? | Yes | |||
| Time to add a new retailer once live? | 3 business days for pre-built integrations |
Requirements: Normalization & Data Model
| Requirement | Alloy.ai | Vendor 2 | Vendor 3 | Vendor 4 |
|---|---|---|---|---|
| Resolves the same product to one master item across every retailer automatically, or do we match it? | Cross-retailer item resolution; automatically where data is available | |||
| Delivers data at the SKU and store/location level, not just national or regional roll-ups? | Yes, SKU-location level, daily (where available) | |||
| How are unknown or new SKUs handled? | Auto-flagged and resolved | |||
| Models locations as a network (store, DC, retailer supply chain)? | Yes | |||
| Normalizes mixed time granularities, currencies, and units of measure? | Yes |
Requirements: Supply Chain & Execution
| Requirement | Alloy.ai | Vendor 2 | Vendor 3 | Vendor 4 |
|---|---|---|---|---|
| Connects past POS into DC, ERP, and 3PL data? | Yes | |||
| OOS & Lost Sales, Phantom Inventory, and New Item Management out of the box? | Yes, out of the box | |||
| DC-level and store-level order recommendations? | Yes (Replenishment), for retailers where the data supports it | |||
| Monitors retailer forecasts and alerts on deviations? | Yes |
Requirements: AI & Agents
| Requirement | Alloy.ai | Vendor 2 | Vendor 3 | Vendor 4 |
|---|---|---|---|---|
| Agentic AI usable today (not roadmap)? What do the agents do? | Yes. Two AI agents are live today: the Replenishment Agent drafts store- and DC-level reorder recommendations, and the Performance Reporting Agent builds recurring reports, both for your team to approve, edit, or reject before anything is sent. More agents are in development. Lens, our conversational AI analyst, goes further: fluent in both CPG analytics and your data, it answers questions, builds visualizations, and acts as a thought partner, helping you frame a problem, find the right data, and decide what to do next. No SQL or dashboards required. | |||
| Review, edit, approve, or reject workflow before an agent sends anything? | Yes, human approves before anything is sent | |||
| Natural-language interface for business users, no SQL? | Yes (Lens) | |||
| MCP server / can our own agents query the platform programmatically? | Yes, MCP server | |||
| Is our data used to train your or any third-party models? | No, never |
Requirements: Data Integrity & Freshness
| Requirement | Alloy.ai | Vendor 2 | Vendor 3 | Vendor 4 |
|---|---|---|---|---|
| How fresh is the data, and is it retailer-direct / unprojected? | Daily (where possible), retailer-direct, unprojected | |||
| Auto-retry failed loads and monitor file status (received / pending / missing)? | Yes, self-healing feeds with auto backfill, integrated tools for monitoring and alerting data status | |||
| Incorporates restatements once provided? | Yes, restatements are auto-imported once available |
Requirements: Forecasting & Analytics
| Requirement | Alloy.ai | Vendor 2 | Vendor 3 | Vendor 4 |
|---|---|---|---|---|
| POS forecasting available? At what granularity and with which models? | Yes, multiple models at SKU-location | |||
| Measures retailer, POS, and internal forecast accuracy? | Yes (3-way reconciliation) | |||
| Global reporting across retailers, retailer-specific KPIs, and ad hoc dashboards? | Yes |
Requirements: Ownership, Security, & Support
| Requirement | Alloy.ai | Vendor 2 | Vendor 3 | Vendor 4 |
|---|---|---|---|---|
| Do we own our data, and can we export it whenever we want? | Yes, you own it; export any time | |||
| Which export destinations are supported? | Data warehouses, BI & analytics tools, ERP & planning tools, and Excel, Slides, etc. | |||
| SOC 2 Type II, encryption at rest and in transit, isolated instance? | Yes to all | |||
| Time to first data live, and full rollout timeline? | 3 business days for first feed; ~8 weeks for 10+ retailers | |||
| Who does the implementation work? | Alloy does the build (white-glove) | |||
| Support model and international coverage? | White-glove, international coverage |
Alloy.ai is a commerce intelligence system for consumer goods brands. Through 500+ integrations with retailers, e-commerce platforms, online marketplaces, social commerce channels, distributors, 3PLs, and ERPs, Alloy brings in the actual sales and inventory data commerce partners report, not estimates, and reconciles it with internal shipment data for one consistent view down to the SKU and location. On that foundation, Alloy delivers the analytics that brands run their business on, and its AI Agents turn what the data shows into replenishment orders and executive reports for humans to review and approve. Sales and Supply Chain teams at brands including Liquid I.V., Crayola, and Valvoline use Alloy to recover lost sales, protect margins, and keep products on the shelf.
Book a Demo
See Alloy.ai
Synchronize execution to eliminate waste, mitigate risk, and capture every revenue opportunity.
Once you submit a demo request form:
- A sales rep will email you within 24 hours to schedule a brief intro call.
- Our team will provide an overview of Alloy.ai and learn more about your data goals and objectives.
- We’ll create and walk you through a live demo, customized to your specific data needs.